v5.2.0c under Xubuntu 8.10 not scheduling reports to email FATAL >>>> error in log

[gazza73]

Sugar v5.2.0c CE
under Xubuntu 8.10

Error trying to schedule a regular Zucker report to to to an email

Runs OK in test Schedule (gets an email to the recipient with the PDF attachment)

But: A HOURLY scheduled report results in no emil being sent.
An error comes up in /var/www/sugarcrm/sugarcrm.log:

Fri Apr 3 17:05:03 2009 [11999] [1] [FATAL] >>>>
Fri Apr 3 17:05:03 2009 [11999] [1] [FATAL] --->>>>

9 pairs of these errors come up, every 60seconds as the Xubuntu CRON Daemon does another pass of the cron.php execution.


We put Xubuntu and 5.2.0c tpgether overnight becuase we were faling to get cron to work on Linux 2.6.24.4 with v5.0.0.c as well (see posts by gazza73 on this yesterday.

Stopping and starting cron
and deleting the /var/www/sugarcrm/cache/modules/Schedulers/php.pid file
makes no difference. The FATAL error occurs as the cron cycle runs past the second time.

Any ideas on this error.

So far, I've tried FreeBSD v6.2 with v4.5.1
Linux 2.6.24.4 with v5.0.0.c (recommendd download from Sugarcrm demo site_
now Xubuntu v8.10 with latest v5.2.0c
and I've yet to see any automated scheduled reports running .......

I must be missing something somewhere......

Help appreciated!

Gaz
Australia

[gazza73]

sugar 520c CE
Xubuntu 8.10


getting more confuse about this with the CE version.

It seems to be running IF I have the firefox/sugar session running on a browser..... Is that a requirement?

I'd like to have this scheduling running by the sugarCRM CE server in the background overnight - not when users are runing a browser session - or even the 'admin' user running a session......

The error persists in the sugarcrm.log file

Fri Apr 3 21:49:03 2009 [31267] [1] [FATAL] --->>>>1
Fri Apr 3 21:49:03 2009 [31267] [1] [FATAL] >>>>


but if I reopen and leave running, the firefox session to sugar, the actual PDF emails do arrive every hour!!!

When the [FATAL] error occurs, I managed to see the following jobs in #ps ax :

This is the job running at the instance of the error:

31266 ? Ss 0:00 /bin/sh -c cd /var/wwww/sugarcrm; php -f cron.php > /dev/null 2>&1
31267 ? R 0:00 php -f cron.php
31268 ? R 0:00 /usr/lib/hal/hal-acl-tool --reconfigure
........


So the error in sugarcrm.log is directly related to the process (31267 in this example) to the executon of the php -f cron.php

Any ideas......? Maybe I just have a bad typo in the cron.php.

But I still don't understand what was stopping the cron producing emails automatically unless I was running a suagr session in a client machine's browser?

[gazza73]

upgrading with the PATCH for v5.2.0f fixed this problem.

Also, the /etc/crontab file needs to have the line:


* * * * * root cd /var/www/sugarcrm && php -f cron.php > /dev/null 2>&1

With Ubuntu, the syntax for setting up the cron command in /etc/crontab is different:

- with the 'root' user mentioned,
- and use of the '&&' characters to run both commands.

Gaz